Materials and methods
We reviewed retrospectively data of consecutive 246 patients with COM who underwent surgical management at our hospital between Jan 2019 to Dec 2020. 173 patients with pre-and postoperative pure tone audiometry data were included in the analysis. Exclusion criteria included patients with revision surgery, adhesive otitis media, and unsuccessful surgical. The surgical procedures were performed by one otology surgeon. The selection of surgical approach is determined by the range of lesion as well as the preoperative status of hearing and facial nerve function. All procedures involving human participants followed the institutional research committee ethical standards in accordance with the 1964 Helsinki declaration and its later amendments. This was retrospective study, thus no informed consent was needed.
Pure tone audiometry was conducted in a double-chamber anechoic room using standard procedures. All patients had pre-operative pure tone audiometry within one week of surgery. Air conduction threshold (AC), bone conduction threshold (BC), and air-bone gap (ABG) were tested at 250, 500, 1000, 2000, and 4000 Hz, respectively. One year following surgery, the same methods were used to examine post-operative pure tone audiometry. At these frequencies, changes in AC, BC, and ABG were calculated between pre- and postoperative examinations. Statistical analysis was performed for each frequency by comparing pre-and postoperative mean values. A positive value indicates improvement and a negative value indicates deterioration.
Based on possible influence factors, these COM patients were retrospectively categorized into three subgroups (Table 1): group A (based on the type of surgical method), tympanoplasty vs. canal wall-down tympanoplasty; group B (based on the type of ossicular replacement), partial ossicular replacement prosthesis (PORP) vs. total ossicular replacement prosthesis (TORP) vs. none; group C (based on the type of tympanic membrane repair material), temporalis fascia vs. tragus cartilage.
Statistical analyses were performed using SPSS 21.0 software. The metering data were expressed as mean ± SD and the significance analysis using t-test. The categorical variables were compared using the Chi-square test or Fisher’s exact test. P < .05 indicates that the difference is statistically significant.
Table 1 Distribution of cases in different groups.